Doug,
So I think the problem is that "HackRF=0" isn't recognized, so you have
the original problem that you had with the source. Basically the driver
selection falls through to the first driver that says "im here" -- which
in this case is a false positive.
Its not an ubuntu vs windows issue, its just that a lot more sdr driver
software is installed in this windows installer than you would typically
have on ubuntu where its installed piecemeal.
I agree that it would be great if it just selected the first available
device and worked. I think in this case, there is a misbehaving driver
that says "im here" when it shouldn’t. That needs to be fixed, but for
now with osmocom source/sink, you do have to be explicit with the device
args.
So for the transmitter, gr-osmosdr sink should expect "hackrf=0". I have
no idea why its dumping that usb error: HACKRF_ERROR_NOT_FOUND
Obviously, hackrf source has seen a lot more testing.
One suggestion I have is to use the SoapyHackRF instead. We had issues
related to the gr-osmosdr hackrf support, and a user contributed
SoapyHackRF, which is an alternative wrapper to the gr-osmosdr hackrf.
To invoke it with the osmocom sink, use the args "soapy=0,driver=hackrf".
soapy=0 tells gr-osmosdr to use the soapy sdr support, and driver=hackrf
tells soapy sdr to use SoapyHackRF.
-josh